Get a Demo
Our Technology

Our cutting-edge voice technology sets a new standard in authentication, offering a level of security and precision unmatched in the industry. Combining advanced biometric analysis with real-time detection, our solution is uniquely designed to provide seamless, highly accurate identity verification that is virtually impossible to spoof. Here’s what makes our technology stand out.

Multi-Dimensional

It measures both the behavioral traits of an individual as well as the impact their physical body creates on sound

Spoof-Proof Voiceprint

Creates a unique spoof-proof ‘Voiceprint’ of the individual, by combining their sound, voice pattern and rhythm

Acoustic Characteristics

Detects inaudible acoustic traits, ensuring the highest Identity Assurance and full confidence in verifying claimed identities

Real-Time Fraud Detection

Instantly identifies and prevents suspicious or fraudulent activity during authentication

State of the Art Tech Architecture
Flexible Architecture
Support for multiple database technologies & operating systems
Flexible Deployments​
Cloud-based SaaS, on-premise and on-device deployments supported
Low Latency
Developed using modern programming techniques on Java, Go, and C++ programming technologies
Assistive AI​
Augmenting Assistive AI technologies to enable secure voice driven commands
API Driven Architecture​
API driven architecture to support existing operational, support and reporting infrastructure
Deployment and Integration

State-of-the-Art Technical architecture ensures the flexibility to meet ​the most demanding requirements in a rapidly changing landscape.

Enterprise Support
ValidSoft's responsive services and support teams bring deep expertise to help ramp fast
Flexible Platform
True multi-tenancy SaaS plus on-prem deployment options. Flexible, scalable, extensible
Pre-integrated
Existing integrations with leading call center tech providers
Open API Platform
Full SDK available. Robust APIs to meet the most sophisticated enterprise requirements
Reach Out
Our Patents

20+ patents granted. More on the way. ValidSoft owns all of its intellectual property

Anonymous Correlation System

For fraud prevention

Incremental Contextual Enrollment

Voice biometric enhancement

Deepfake Audio Detection

Recognition of special voice characteristics

VALid

Out-of-band authentication method

Dual-SSID

Wireless network authentication method

VALid-POS

Payment card and mobile device location correlation

VALid-CFU

Fraudulent call-forwarding detection type-1

Nemesis-PDT

App-based fraud prevention

VALid-CFA

Fraudulent call-forwarding detection type-2

VALid-SSD

Fraudulent SIM-swap detection

Screen Scrape & Replay Detection

Dual-channel MTB Prevention

Detecting Call Redirection

Fraud prevention and identity confirmation

AI in Cybersecurity Award

ValidSoft’s Voice Verity® was recently rewarded an AI in Cybersecurity Innovation Award by TMC.net

“In an era where cyber threats are becoming increasingly sophisticated, the importance of innovative security solutions has never been greater. TMCnet’s AI in Cybersecurity Innovation Award shines a spotlight on those companies and technologies that are leading the charge in defending digital landscapes through artificial intelligence.

“We are thrilled to recognize the companies leading the charge in utilizing cutting-edge AI technologies to fortify cyber defenses, said Rich Tehrani, CEO, TMC (News – Alert). “These award winners have demonstrated remarkable vision and technical excellence, providing vital tools to safeguard organizations and individuals from cyber threats.” Read full article here.

Unique Approach to Data Privacy

We understand the value of your customer and take the protection of their data seriously

  • No other Voice Biometrics provider, or location-based services provider has attained a Privacy Seal from EU
  • ValidSoft has 4X European Privacy Seals (EuroPriSe®) related to its solutions: Voice Biometrics; Sim Swap; Proximity Correlation (LBS); Zero Latency Correlation
  • ValidSoft can confer these Privacy Seals on its partners
  • ValidSoft’s EuroPriSe® privacy seals ensure 100% compliance with EU GDPR and all other leading Data Protection and Data Privacy laws (incl. e.g., HIPAA, Digital Identity Guidelines, Vectors of Trust, Federal Identity Program Guidelines, etc.)
  • ValidSoft has leveraged these European Privacy Seals to demonstrate 100% compliance with leading global countries including: US, Brazil, Mexico, China, Russia, Australia, South Africa, Turkey, Indonesia and others
How it works

Explore the seamless process of voice biometric authentication, from voice capture to secure identity verification

loading='lazy'
User speaks
loading='lazy'
Raw audio received
loading='lazy'
Key features extracted
loading='lazy'
DNN inference
loading='lazy'
Voiceprint created (no PII)
See How ValidSoft Can Move Your Business Beyond Passwords

Get access to Validsoft demo portal

Request Demo
Decor Decor
Industry-leading Features

Some of the Features of our Engine

Passive Authentication
Passive, or, Text-Independent Authentication via natural speech or conversation. Reduces security questions.
Conversational Authentication
Voice biometrics (speaker recognition) combines with speech recognition to verify what was said, what was requested.
Precision Voice Biometrics™
Active authentication combined with randomly generated one-time passcode or one-time sms to a known device.
Multi-voiceprint Management
Context-based support of multiple voice prints per user. i.e. one for text independent one for tex-dependent authentication.
Voiceprints Clustering
Cluster voiceprints into groups with common purposes, i.e. identify a speaker within a group of enrolled users.
Offline Ops Management
Identification, verification, cross checking are done offline. Helps with fraud prevention (multiple false identities).
Synthetic Voice Detection
Protects integrity of voice based system, detects bot calls, supplements Data Mining applications for Government agencies.
Replay Detection
The system analyses the audio provided to detect whether it is a recording to protect against fraud via replay.
Multi-language Support
Fully language agnostic algorithms to support multiple language as standard.
Utterance Verification
Ensures accuracy of acoustic characteristics. Ensures better enrollments and verifications.
Liveness Support
Liveness Validation ensures a “Live” user is speaking during voice enrollment or verification.
Watchlist Processing
Identification of known fraudsters through a federated blacklist or the ID of premium users/clients.
Whitelisting
The system provides the capability of defining a list of voiceprints of known or legitimate users for use in authentication.
Speaker Separation
Identifies different voices in the same audio. Useful in call centers (one audio file has both client and agent voices).
Blacklisting
The system provides the ability to define a list of voiceprints as a blacklist for checking during authentications.
Future-Proofing
As voices change over time, Future Proofing automatically updates the voice prints based on pre-defined criteria. This helps ensure performance.
Frequently Asked Questions
What areas of ValidSoft’s product can be customized and to what extent?

ValidSoft owns its entire software stack and theoretically everything can be customized inline with client requirements.

What infrastructure requirements does ValidSoft have?

Minimally even a single low power host can operate our solution offering single figure concurrent TPS. We suggest two modern cores, 4GB, 20GB storage in these scenarios

We support virtual servers and our solution can be deliver and run in containerized form.

Are there any losses in functionality with ValidSoft’s cloud-based solutions?

All product functionality between the host/cloud-based service are the same. There may be some difference in the alerting or logging technologies used, as we leverage some native service provided by the IaaS provider. However, this is nothing intrinsic to the functionality of the product or platform.

Does the hosted/cloud-based solution need to be hosted in ValidSoft’s data center, or can it be hosted in a third-party environment?

Validsoft’s solution is containerized and, whilst we encourage a VPC based approach, the solution is ultimately portable between traditional DC infrastructure, virtualized solutions and multiple cloud landing zones. Customers may specify their requirements.

Does ValidSoft have any compliance related certifications?

Validsoft retain ISO27001 certification and operate ITSM to this standard in all respects.

ValidSoft’s solutions are built using a “Privacy by Design” approach that places an emphasis on data privacy and protection and ensures compliance with all regulatory regimes including, but not limited to, GDPR, CCPA, HIPAA, BIPA, LGPD and Vectors of Trust.

ValidSoft’s voice biometric solution when used in accordance with the Terms & Conditions is compliant with the following data privacy acts, although this is not an exhaustive list.

  • General Data Protection Regulation (GDPR) (EU)
  • General Data Protection Law (LGPD) (Brazil)
  • California Consumer Privacy Act (CCPA) (US)
  • Biometric Information Privacy Act – Illinois (BIPA) (US)
  • Health Insurance Portability and Accountability Act (HIPAA) (US)
  • NIST Digital Identity Guidelines (US)
  • IETF Vectors of Trust (VoT)
  • Federal Identity Program Guidelines (FIP) (Canada)

ValidSoft has also leveraged its European Privacy Seals to demonstrate 100% compliance with other leading global jurisdictions.

What is the minimum amount of net audio required for enrollment?

Current Voice Biometrics technology is state of the Art and does not stipulate a minimum audio length to perform an enrollment. Our recommendation is to capture enough audio for a proper quality enrollment. Depending on use case, environment and infrastructure, minimum audio recommendation in call centers for a strong enrollment is between 10 to 20 seconds for Passive and 6-9 seconds for a complete Active/Dynamic enrollment. Enrollments with less audio are possible, but voice print quality, and therefore authentication accuracy, can be impacted. Equally, longer enrolment durations will result in a higher voice print quality and better authentication success. Our system is also able to add additional audio to existing enrollments, over multiple sessions/calls based on pre-defined business logic and variable thresholds.

What is the minimum amount of net audio required for authentication?

Per the above answer concerning enrollment, there is no minimum audio length requirement for authentication. We do recommend collecting at least a minimum of 2-3 seconds of audio in Passive and a single voice phrase of 1-3 seconds for Active. Again, having more audio available for Passive – even just 5-6 seconds – can drastically improve the authentication results and accuracy.

Is ValidSoft’s voice biometric technology designed for a specific type of device?

In short, no. ValidSoft’s voice biometric systems are designed to work with a variety of devices and conditions. When an audio sample is acquired, it is inevitably affected by nuisance attributes (attributes not contributing to specifying the individual) such as the type of microphone, the background noise, the transmission channels, etc. Normalization techniques are applied to compensate for these effects.

How does ValidSoft adhere to recommended guidelines (NIST, or otherwise) for FAA and FRR?

ValidSoft recommends at least one tuning phase during the actual pilot/production setup under controlled conditions with a pilot group of users. We recommend using 300+ individuals to perform at least one enrollment and one verification call on multiple days; the more the better. These calls should be done under controlled conditions where no impostor tests or other negative tests are performed. The audio can then be used to create enrollment and evaluation sets with significant true and false trials similar to how NIST evaluation datasets are designed. Thresholds are adjustable and typically operate at a 1-5% difference in FAA/FRR based on customer’s preferences surrounding security and user experience.

Does ValidSoft hold any patents/intellectual property for its technology?

ValidSoft have been granted 22 Patents in Speech- and Authentication- Technologies with more pending.

Want to Know More? Reach Out!
Learn how ValidSoft can revolutionize your business with seamless voice authentication, deepfake protection, and real-time fraud prevention—book your demo today and elevate your customer security experience.
Icon
Our relationship with ValidSoft initially began as a straightforward supplier arrangement. However, as our needs evolved and diverged from their standard offerings, ValidSoft showcased great flexibility and collaboration. They not only addressed our immediate requirements but also committed to continuous innovation, providing regular updates and enhancements. This proactive approach has elevated our relationship into a true partnership, where both Invoca and ValidSoft are mutually benefiting from the leading edge biometrics solution now in place.
Beatrice Casanova
Vice President Product Management

    Please fill in the field
    Please fill in the field
    Incorrect email format
    Please fill in the field
    Please fill in the field
    ValidSoft is committed to protecting and respecting your privacy, and we’ll only use your personal information to administer your account and to provide the products and services you requested from us. From time to time, we would like to contact you about our products and services, as well as other content that may be of interest to you. If you consent to us contacting you for this purpose, please tick below to say how you would like us to contact you:
    Get a Demo